Processors - Page 5

GIGABYTE has just achieved the CPU frequency world record for AMD's new Ryzen 7 9800X3D processor, using its X870 AORUS Tachyon motherboard, pushing the 9800X3D past the heights of 7.3GHz.

At a recent product launch event in Shanghai, China, GIGABYTE's in-house overclocker -- HiCookie from Taiwan -- put his skills to the test yet again, and walked away with the frequency world record for the 9800X3D on the X870 AORUS Tachyon motherboard.

AMD's new Ryzen 7 9800X3D processor has 8 cores and 16 threads at up to 5.2GHz with 96MB of cache and a 120W TDP, HiCookie overclocked the CPU under LN2 cooling, hitting 7313MHz (7.31GHz) and was the first one to breach the 7.3GHz barrier as the previous record holder reached (just) 7.2GHz.

Apple unveiled its new iPhone 17 smartphones last week powered by its new A19 Pro and A19 processors, fabbed at TSMC on its new 3nm process node, and now we have some comparisons between the A19 Pro and A19 chips. Check it out:

We don't yet know how many transistors are in the new A19 series processors, but we know their codenames are "Thera" and "Tilos" with both the A19 Pro and A19 featuring the same 6-core CPU with identical clock speeds (2 x 4.26GHz + 4 x 2.60GHz). The L2 shared cache amount is different, with the A19 Pro featuring Big 16MB + Small 6MB while the A19 sports Big 8MB and Small 4MB.

The GPU count on the A19 Pro is at 96 EUs (Execution Units) while the A19 has 80 EUs, while GPU ALUs on the A19 Pro sit at 768 ALUs while the A19 has 640 ALUs. Apple's new A19 Pro chip has 6 GPU clusters while the non-Pro variant has 5 GPU clusters.

Qualcomm is all systems go to unveil its next-gen Snapdragon 8 Elite Gen 5 mobile processor next week, with the company outing it a little early, and explaining why it's changed the name of its upcoming SoC.

The company admits that it might look like it's skipped a few generations, but Qualcomm says the truth is simpler -- and more powerful -- as the new Snapdragon 8 Elite Gen 5 marks the fifth generation of its premium 8-series platforms since Qualcomm shifted to its single-digit naming and visual identity years ago. Qualcomm adds "think of it like this: Gen 5 isn't just a number. It's a signal that this platform leads the family forward".

Qualcomm launched its Snapdragon 8 Elite last year, with the expected release of the next-gen Snapdragon 8 Elite Gen 2, but the shift directly into the Snapdragon 8 Elite Gen 5 will definitely have users' heads spinning. Qualcomm can't quite nail down where they are with their naming scheme, which is why they've had to pre-announce the new naming before announcing the chip itself.

We've got the full specs on the semi-custom APU made by AMD and fabbed by TSMC for Sony's next-gen PlayStation 6 console, which should easily handle 4K 60FPS gaming without AI upscaling, and up to 4K 120FPS in thousands of games.

In his latest video leaker Moore's Law is Dead has provided the full leaked specs for the PS6 console and its AMD Orion APU, after months of leaks on the Orion APU, now we've got the nitty gritty. Inside of the PS6 will be the Orion APU with a 280mm2 monolithic die fabbed on TSMC's new 3nm process node, with up to 160W TDP.

The monolithic die and 160W TDP are highlights of AMD's semi-custom Orion APU for the PlayStation 6, as AMD is also making a semi-custom "Magnus" APU for Microsoft and its next-gen Xbox console, but Magnus is NOT a monolithic die but a chiplet-based die, with a higher TDP. The higher TDP = more cooling = more expensive. The chiplet Magnus APU die = more cost, more cooling = more expensive... meaning Sony could price its next-gen PS6 console at the same price of the PS5 Pro.

We're all aware that Intel changed the naming of its Core processors a while back now, but a new model of CPU has emerged that confusingly reverts to the old convention.

Yes, a Core i5 processor just arrived, albeit this is a strange release from Intel in that it's from the Comet Lake generation (remember those chips?).

Specifically, this is the Core i5-110, which rather oddly mixes the old naming scheme (i5) with the new one (110). This CPU is actually an addition to the Comet Lake-S range, and as VideoCardz points out, it's essentially a refreshed Core i5-10400.

When it comes to all things tech, there's nothing bigger than CES, which takes place annually in January. With CES 2026 just a few months away, AMD has announced that its CEO, Dr. Lisa Su, will deliver the opening keynote on January 5, 2026, in the Palazzo Ballroom at the Venetian Hotel in Las Vegas.

According to the announcement, the keynote will focus on AI solutions and AMD's portfolio of CPUs, GPUs, adaptive computing, and AI software. This includes data center products, such as the company's EPYC processors and Instinct GPUs, as well as Ryzen processors and Radeon graphics for "AI PCs and gaming." Although it makes perfect sense for AI to be the focus, it's great to see Ryzen and Radeon get a mention in the announcement, as it suggests that we could be in for a few big announcements from AMD at the show.

Hosting the opening keynote doesn't mean that AMD is set to unveil its next-generation Ryzen CPUs and Radeon GPUs. Still, we'll probably hear about a range of new devices and solutions, or even a teaser for where Zen 6 or next-gen RDNA 5 technology is headed. Perhaps we'll hear about an RDNA 4 refresh? Time will tell, but Dr. Lisa Su's comments do state that CES is the "premier stage" for announcements and breakthroughs.

Intel's Arrow Lake Refresh processors were rumored to be set for arrival later this year - but that's not going to happen.

Intel has confirmed that there will indeed be an Arrow Lake Refresh, as the grapevine had insisted (more recently, after falling out with the idea that it might have been canned). However, this refreshed range of CPUs won't be out until 2026.

Tom's Hardware reports that Intel's Corporate VP of Investor Relations, John Pitzer, made the revelation at a Goldman Sachs Technology conference.

Intel's Application Optimization (APO) optimizes thread scheduling for select CPUs to ensure efficient resource allocation. When paired with games with Intel APO profiles, the direct result is improved performance across average frame rates and the all-important 1% lows. Recently, Intel announced that it has expanded APO support to include 15 new games, covering a wide range of genres.

As APO is available for Intel's current desktop lineup of Intel Core Ultra 200 Series CPUs, the company has provided internal benchmarks showing that with APO enabled, game performance can be improved by up to 14%. According to the chart, you can see that when APO is paired with the flagship Intel Core Ultra 9 285K and a GeForce RTX 5090 GPU, performance in the ray-tracing-heavy Metro Exodus Enhanced Edition improves by 14%.

Of course, this is on the extreme end of the scale, with popular multiplayer title World of Warships only seeing a 3% increase in performance and single-player sci-fi horror title The Callisto Protocol seeing a 4% increase. On the plus side, the APO improvements also improve 1% low performance, which, for World of Warships, improves by a respectable 9%.

Intel's mid-range Core i5-14600K processor has dropped to a new low price of $149 on Newegg, including a free copy of Battlefield 6, making it an insanely great CPU for the money.

Intel is giving away free copies of Battlefield 6 with its Core Ultra + Core CPUs as part of its "Gamer Days 2025" deals, with the Core i5-14600K dropping to $189 on Newegg but when mixed with a $40 off promo code, drops to an incredibly awesome $149.

The Core i5-14600K processor ships with 14 cores and 20 threads of CPU performance for the LGA 1700 platform, with more than enough processing power to run Battlefield 6 at high 120FPS and beyond. If you consider that Battlefield 6 costs $69 on its own, it means that the Intel Core i5-14600K drops down to only $89 on its own... a bloody impressive deal.

AMD will reportedly be using TSMC's new 2nm and 3nm process nodes for the CCD and IOD, respectively, on its next-gen Zen 6-based Ryzen CPUs, solidifying rumors over the last few months.

In a new post from leaker "Kepler_L2" we're hearing that AMD's next-gen Zen 6-based Ryzen CPUs will use TSMC's bleeding-edge N2P process node for its CCDs and the N3P process node for the IOD, with the 2nm ramp from TSMC expected to begin in Q3 2026.

AMD has already confirmed that its next-gen Venice CCDs featuring new Zen 6 cores will be fabbed on TSMC's new 2nm (N2P) process node, while the IOD will be fabbed on TSMC's new 3nm (N3P) process node. Right now, Zen 5 chips are fabbed on TSMC's 4nm process node for the CCDs and 6nm for the IOD.

Intel's Nova Lake flagship may indeed be a 52-core processor as previously rumored, as another piece of evidence has emerged indicating this could be the case.

VideoCardz aired a shipping manifest which contains multiple mentions of 'NVL-S' chips - meaning Nova Lake-S desktop processors - with 28C and 52C flavors (28-cores and 52-cores).

As past speculation has detailed, the flagship Nova Lake CPU is theorized to run with 16 performance cores, 32 efficiency, and 4 low-power cores for that total of 52. The 28-core model mentioned is the Core Ultra 5 (i5) workhorse with 8 performance cores plus 16 efficiency and 4 low-power.

A pair of new Zen 5 processors are seemingly imminent from AMD, or at least these chips have been listed by an online retailer.

The CPUs in question are the Ryzen 9700F (8-core) and 9500F (6-core) processors as flagged by leaker Harukaze5719 on X (via VideoCardz).

As you're likely aware, the 'F' suffix denotes that these chips don't have integrated graphics, and there have been leaks about both these products in recent times.

Intel has been suffering against AMD for many years now in the CPU business, but the tide is slowly turning... while a newly-filed patent from Intel teases software-defined Super Cores, which would see smaller cores working together as a bigger Super Core to boost single-threaded performance.

Normally, we'd see larger cores used in a CPU but they've got their limits, with a single huge core having diminishing returns, with Intel's new patent "EP4579444A1" showing how Intel wants to overcome these issues through an SDC aka "Software Defined Super Cores".

The idea behind SDC is that Intel would use more cores than a single large core, fusing them together -- virtually -- when required. For example, two smaller cores could work together instead of a larger core dividing the workload, which would boost single-threaded performance considerably.

Qualcomm's next-gen Snapdragon 8 Elite Gen2 might shift into high marketing gear, with rumors swirling that it will be dubbed the Snapdragon 8 Elite Gen5, and that we'll see huge 4.61GHz clock speeds and 50% more performance than its predecessor.

In new rumors from Weibo tipster Digital Chat Station, we're hearing new information on Qualcomm's purported Snapdragon 8 Elite Gen5 processor, which was reported tested on Samsung's unreleased Galaxy S26 Edge with a higher-binned part as its performance cores on Geekbench 6 were a huge 4.74GHz, instead of the regular 4.61GHz in previous rumors.

Qualcomm seems to be going with the same 2 + 6 cluster of CPU cores like the Snapdragon 8 Elite, with the Performance cores at a default of 4.61GHz and a tweaked (through higher-binned silicon) of 4.74GHz, with the Efficiency cores at 3.63GHz. On the GPU side, Qualcomm is using the Adreno 840 with the leaker saying that we'll see default clocks of 1.20GHz, which is 100MHz more than the 1.10GHz on Adreno 830.

Intel has admitted it dropped the ball with desktop processors and Arrow Lake CPUs, hoping to turn things around with its next-generation Nova Lake CPUs in 2026.

Intel Chief Financial Officer David Zinsner spoke at the Deutsche Bank 2025 Technology Conference, talking about Intel's upcoming products including desktops and servers, with some comments on consumer CPUs.

He mentioned Intel's high-end CPU positioning and that its Arrow Lake chips had lower gaming performance than their predecessors, and that AMD had picked up aggressively in the same space with its Zen 5-based Ryzen CPUs. AMD has done a phenomenal job with its X3D processors as well, with the super-popular Ryzen 7 9800X3D being a kick-ass gaming CPU.

ASRock has released a new BIOS for its AM5 800 series motherboards, and the hope is that this could be an end to Ryzen 9000 X3D processor burnouts with these boards.

VideoCardz noticed that Uniko's Hardware flagged up the new BIOS in a post on X which you can see above.

ASRock's new BIOS is version 3.40 and it comes with two items in the release notes, saying that it improves "memory compatibility and system stability" and also enhances "CPU operating stability".

Intel has provided us with some rather exciting news at the Hot Chips 2025 event, with details on its next-gen Clearwater Forest Xeon CPUs: up to 288 cores, made on Intel's in-house 18A process node and on American soil, with Foveros Direct 3D construction.

The new Intel Clearwater Forest Xeon CPUs will have only Efficient cores (E-Cores) and no Performance cores (P-Cores) with up to 288 cores, with the company moving the package on its next-gen Clearwater Forest Xeon processors into chiplets, with 12 compute chiplets fabbed on Intel 18A, above three base tiles fabbed on Intel 3, and two IO chiplets on Intel 7.

Intel then links this all through its in-house Foveros Direct 3D stacking and EMIB technologies, with a monolithic mesh fabric that connects resources across the package. Clearwater Forest Xeon CPUs are designed and fully built in the US, versus processors being fabbed at TSMC in Taiwan, marking a huge milestone for Intel.

Earlier this year, we began hearing about an alarming number of AMD AM5 X3D CPUs burning out and failing - namely, the current PC gaming champ, the Ryzen 7 9800X3D. And although the failures have covered a wide range of motherboard manufacturers, it does seem that a large number of cases are tied to ASRock's motherboard line-up.

There have been several investigations into the issue, from AMD, ASRock, and independent media like Gamers Nexus, to determine the cause of the problem. The "9800X3D Failures/Deaths Megathread" on Reddit has been tracking this problem for around six months now, with the fix or preventative measure for those using a 9800X3D CPU being to ensure that they update their motherboard to the latest BIOS.

In fact, some of the CPU damage that we've been seeing has been linked to overly aggressive Precision Boost Overdrive settings. In a new interview with Quasar Zone, AMD's David McAfee and Travis Kirsch have provided an updated statement on this issue, with the company seemingly blaming motherboard manufacturers for being too lax or taking advantage of its openness to overclocking and custom Precision Boost Overdrive settings.

Samsung Electronics could partner with Intel now that it has the backing of President Trump and the US government, in a move to solidify US semiconductor manufacturing against TSMC in Taiwan.

The US government recently secured a 10% equity stake in US chipmaker Intel, with Samsung reportedly looking into a "strategic partnership" with Intel, in order to secure itself a better trade deal with the Trump administration according to new reports from the Taiwan Economic Daily.

Samsung is reportedly looking at partnerships with American companies -- recently securing a $16.5 billion foundry contract with Tesla to fab its new AI6 processors -- in order to keep the Trump administration happy, ensuring its American operations weren't hit with huge tariffs.

Intel's next-gen Panther Lake CPUs aren't out on the market yet, but the first motherboard supporting the new Panther Lake processors has been spotted, a new Mini-ITX industrial motherboard. Check it out:

There are a few interesting things here with this new Mini-ITX industrial motherboard, first of which is that it's from the LanParty motherboard maker DFI, and secondly it features an Intel Panther Lake-H processor in 25W TDP form, made on Intel's new in-house 18A process node.

Intel's new Panther Lake-H CPU architecture features new Cougar Cove P-Cores, Darkmont E-Cores, and next-gen Intel Xe3 graphics. Panther Lake chips will sport up to 16 cores (a mix of P-Cores, E-Cores, and LP-E Cores) with up to 12 GPU cores based on the new Xe3 architecture.